Process Philosophy and Christian Thought
by Delwin Brown, Ralph James, Gene Reeves (eds.)

Chapter 11: The Formally Possible Doctrines of God by Charles Hartshorne

Abridgment of ‘The Formally Possible Doctrines from Man’s Vision of God." Copyright, 1941 by Charles Hartshorne. Reprinted by permission of Harper & Row, Publishers and Charles Hartshorne. Charles Hartshorne received his degree from Harvard University, where he was Whitehead’s assistant. He has taught at Harvard, Chicago, and Emory Universities and he is now Ashbel Professor of Philosophy at the University of Texas.

For nearly two thousand years European theology staked its fortunes upon a certain conception of divinity. In spite of the seeming variety of doctrines, one basic principle was accepted by almost all philosophical theists. Only in the last few decades has a genuinely alternative type of theology been at all widely considered — so unobtrusively, however, that many opponents of theism, even some of the most distinguished, are still fighting the older conception exclusively, convinced that if they can dispose of it the theological question will be settled. And many of those who find the idea of a godless universe incredible suppose that it is to traditional theology that they must turn. Both parties are mistaken. Today the theistic question, like many others, is a definitely new one. Many of the old controversies, in their old forms, are antiquated.

As traditional theology was a relatively well defined system, the same in certain basic respects — despite all sorts of philosophical and ecclesiastical differences — in Augustine, Thomas Aquinas, Maimonides, Leibniz, Calvin, Immanuel Kant, and some schools of Hindu thought, so the new theology which many be contrasted with the old is found more or less fully and consistently represented in thinkers as far apart as William James, . . . Henri Bergson, F. R. Tennant, . . . A. N. Whitehead, . . . Nicholas Berdyaev, . . . and in numerous others of every brand of Protestantism, besides a few . . . Roman Catholics. I have also heard a clear statement of some aspects of it from a leading Hindu thinker, Radhakamal Mukerjee. Of course, there are interesting differences between these theologians, just as there were between Bonaventura and Calvin; and in some writers now, as of old, the logical implications are more adequately and rigorously worked out than in others. But there are some fundamental points of agreement which are rapidly becoming standard among non-Roman Catholic theologians.

To be aware of these points of convergence is essential to a liberal education today. They are as characteristic of our time as relativity physics and logical positivism are, or as medieval theology was of the thirteenth century. Ideas which until about fifty years ago were almost wholly neglected, never clearly worked out and systematized, and perhaps passed over for centuries with scarcely a mention, are now to be met in scores of theological works and in philosophical works that deal carefully with theology. The time seems at hand for attempts to state clearly the revolution of thought through which we have been passing.

What is the "new" doctrine? We shall see presently that it must be an expression of one of the three and only three formally possible views (including atheism and positivism as special cases of one of the three) regarding the supreme being, and that there are reasons for characterizing the new view as that one of the three which is related to the main line of the tradition as a carefully qualified assertion is to an unqualified one, and related to atheism (and certain heretical extremes of theism) as to an unqualified denial. In other words, it is related to the two other possible views as a "higher synthesis" to its "thesis" and "antithesis," as embraced and corrected in a "higher unity," or as a balanced whole truth to its two contrasting half-truths. From this standpoint traditional atheism and traditional theism are two sides of the same error, one of the most characteristic errors of human thought.

An immediate objection to the suggestion of a new idea of God will doubtless be that the term God as defined by usage properly means the God of the religious tradition. But we must distinguish, in the tradition, between religion and theology. Granting that "God" is a religious term, and that theology attempted to describe the object of religious devotion, it is one of the principal questions at issue whether or not this attempt was wholly successful. It is a belief of many today that the "new" theology is more, not less, religious than the old,¹ at least if religion means "devoted love for a being regarded as superlatively worthy of love," which is the Christian conception and to some extent the conception of the higher religions generally.

Of course theologians do not now regard as worthless and merely wrong the entire vast structure of historic theology, any more than Einstein so regards Newton’s physics — to use an analogy which could easily be pressed too far, but whose value could also be underestimated. What is now being done is to distinguish two strands in the theological tradition which were not clearly held apart in the past, and to argue that they are not only distinguishable, but so related that only one of them can be true, and so related also that which one, if either, is true can be ascertained from the logical relations between the two strands alone, since one of the strands is incompatible alike with the assertion and the denial of the other, and hence, by recognized logical principles, is incompatible with itself and necessarily false. It is somewhat — to use another imperfect analogy — like the discovery in geometry of the independence of the parallel postulate from the other assumptions of Euclid; though in the theological case it is not really independence but inconsistency which is involved. Thus it is not a question of the logical possibility, merely, of what might be called a "non-Euclidean theology," but of its logical necessity, at least if there is to be any theology at all. (Unfortunately, there is no individual name which can conveniently serve as the theological parallel to Euclid; but Philo, a Jewish scholar of the first century, might be taken as the first man to give relatively complete expression to the postulate in question, and so we might speak of the current doctrine as non-Philonian theology, in a sense in which Aquinas, Spinoza, Royce, and orthodox Hinduism are all Philonian.²)

The "strand" which theologians, on the whole, still propose to retain, and which is alone self-consistent, as judged by its relations to the other strand, is the popularly familiar definition of God as everlasting, all-controlling, all-knowing, and ethically good or ‘holy" to the highest possible degree. It may seem that this is just traditional theology and must involve the whole time-hallowed system. The extraordinary fact is that this has been found not to be the case. None of the older theologians (unless the neglected — and persecuted — Socinians, and the neglected Jew Gersonides, in the sixteenth and fourteenth centuries respectively, be exceptions) were content with this popular definition of God and the consequences which genuinely follow from it. They invariably adopted other conceptions as even more fundamental; and rather than attempt seriously to deduce these other conceptions from the popular definition, they treated the latter as a more or less dangerously loose or anthropomorphic equivalent of the more fundamental definition. This more fundamental definition turns upon such terms as perfection, infinity, absoluteness, self-dependence, pure actuality, immutability. God, for all the church writers, and for many others, including Spinoza, was the "absolutely infinite," the altogether maximal, supreme, or perfect, being. All his properties, including the popular religious ones so far as philosophically valid, were to be deduced from this absoluteness or perfection, as is so beautifully explained by Thomas Aquinas. . . .

If theology is capable of rejuvenation, its hope lies, I believe, in a re-examination of the idea of infinity or perfection. Perhaps this idea is ambiguous, perhaps there is a sense in which God should be conceived as perfect, another sense in which perfection cannot apply to God, because (it may be) this sense involves an absurdity or, in other words, is really nonsense. Perhaps God is perfect in whatever ways perfection can really be conceived; but some among the traditional theological ways of trying to conceive perfection are capable of producing only pseudo-concepts devoid of consistent meaning.

To discuss God is, by almost universal usage, to discuss some manner of "supreme" or "highest" or "best" individual (or superindividual) being. As a minimal definition, God is an entity somehow superior to other entities. Now such superiority may be merely with respect to other actual entities, or with respect to all entities whether actual or possible. The second or more complete superiority seems to give the appropriate meaning of "perfection," and was defined long ago by Anselm in his description of God as "that than which none greater can be conceived." This definition presupposes only the ideas of something ("that"), greater or more or better (more in value) than, negation or none, and the conceivable or possible, and these ideas are secular as well as religious. Indeed, no ideas are more elementary and unavoidable in philosophy; hence it is clear that religion and philosophy can and must meet on common ground, provided the Anselmian definition successfully defines the religious object. But before we can decide whether the secular terms employed can apply to the God of religion we must be clear as to what the terms mean. Astonishingly enough, the simple phrase "none greater" involves two major equivocations, not indeed as Anselm used the phrase, but as it might reasonably be used, even though the possibility of such usage seems not to have been clearly seen by Anselm or anyone else. The neglected usages constitute, together with Anselm’s usage, a complete set of possible meanings of "perfect being," choice between which meanings is the theistic problem, a problem not fully stated until the neglected meanings are made explicit.

"None" may mean "no entity other than that (the being said to be perfect) as it actually is," or it may mean "no entity other than that as it either is or else could be or become." According to the first meaning (which follows automatically if one assumes that the perfect can have no potential states — an assumption not deducible from the mere idea of "none greater," because of the latter’s equivocal connotation) the perfect is unsurpassable in conception or possibility even by itself; according to the second meaning it is unsurpassable except by itself. The first or absolute unsurpassability can be called absolute perfection, the second may be called relative perfection. (We shall see in the appendix to this chapter, and the reader may have noted, that there is still a third possibility, though apparently it is of no great importance.)

"Greater" has as many meanings as there are dimensions or respects of more and less (or better and worse). But from a purely formal point of view (important because it is exact and non-controversial) there are just three possibilities, two positive and one negative. By "greater" we may mean, "in some (but not all) respects" (say in size or in ethical goodness); or we may mean, "in all respects whatever"; while the joint negative of these two, "in no respect," gives the third possibility.

Combining the two meanings of "none" with the three meanings of "greater" we derive seven possible cases, only one of which is the unequivocal negation of "none greater," or of "unsurpassability even by the conceivable." Thus it is proved that the question, Is there a perfect being? is six distinct questions rather than one. Has anyone a right to assure us, in advance of exploration of the other five, that the Anselmian (unconscious) selection of one among the six — as the faithful rendering either of the religious question or of the most fruitful philosophical one — is safely established by the fact that the choice has been repeated no less unconsciously by multitudes of theologians? If anyone asserts this, I must doubt his understanding of the elementary requirements of good reasoning.

The seven cases can be arranged, in several different ways, into three main groups. The following of the possible triadic arrangements seems the most useful:

Group–Symbol-Case-Symbol-Interpretation

I------(A)----1----A----Absolute perfection in all respects.

II-----(AX)---2----AR---Absolute perfection in some respects,

relative perfection in all others.

--------------3----ARI--Absolute perfection, relative perfection, and "imperfection" (neither absolute nor relative perfection), each in some respects.

--------------4----Al---Absolute perfection in some respects, imperfection in all others.

III-----(X)---5----R----Absolute perfection in no respects,

relative in all.

--------------6----RI---Absolute perfection in no respects,

relative in some, imperfection in the

others.

--------------7----I----Absolute perfection in no respects,

imperfection in all.

Explanation of Symbols: A stands for absolute perfection, R for relative perfection, I for the joint negative of A and R, X for the negative of A (and thus for the disjunction of R and I), and (A) or (X) for the factors occurring throughout a group.

Note: It will be shown in the appendix to this chapter that imperfection can he subdivided into two possible forms, making fifteen cases in all, though the additional eight seem of little importance despite the fact that all eight express modes of unsurpassability, and so of perfection in the most general sense!

In a different mode of presentation we have:

Group----I------II----------III

A in----all----some---------no respects

--------(A)----(AX)---------(X)

Case-----1---2---3----4---5---6-----7

---------A---AR–-ARI--Al--R---RI----I

Note: It might be thought that God’s "supremacy’ requires not only that he cannot conceivably be surpassed, but that he cannot even be equaled. Anyone who wishes to experiment with this conception of the unrivaled as well as unsurpassed is of course at liberty to do so. My reason for neglecting the concept — which might be called "incomparability" — is that I agree with the usual verdict of theologians that the unsurpassable is bound to be unique, so that if superiority is out of the question, equality is also. If good reason for doubting this verdict can be found, then "incomparability’ should be substituted, at least experimentally, for "unsurpassability" in the definition of perfection.

So far as I know, this is the only rigorous formal classification (which as formal and a mere classification is beyond intelligent controversy) of possible doctrines about God — except mere dichotomies (e.g., God is or is not eternal, one with all reality, etc.), which are never very helpful because only one of the two classes has positive content. Yet, though formal, the classification is relevant to religion, if religion believes in an unsurpassable being. And it certainly is relevant to philosophy; for the seven cases (as formal possibilities) follow automatically from concepts which philosophy is bound to use.

At least the classification serves this purpose: it shows how hopelessly ambiguous are phrases like "perfect being," "finite God," "absolute," and the like. Six of the seven cases come under the phrase, "perfect being," if perfection means unsurpassability. At least four are compatible with the description, "finite." Four are definitely included in the class of "absolute" beings. Yet within each classification the differences are at least as important as the resemblances, indeed much more so. For it can be shown that the difference between absolute perfection in all, in some, and in no respects is the crucial difference, and yet it is neglected by all the concepts mentioned and by most generally current ones. . . .

Take, for example, the term pantheism. By any usual definition of this term, it should be possible to give a plausible interpretation of all seven of our cases as conforming to the definition. Thus pantheism means literally almost anything you please, and so nearly nothing. That is probably the chief reason for its popularity as a label for opponents. And it ought to be clear that to say, "God is the all," means whatever one’s view of the all implies, perhaps nothing definite whatever, for offhand we have no clear notion of the all.

It is impossible to think effectively about seven possibilities at once. We think best in threes. As has been shown, the seven possibilities fall logically into three groups. God, if he exists, is absolutely (not relatively) perfect in all, in some, or in no respects. The usual view has been the first. Atheism is a special case of the third, in which man or some wholly imperfect thing is regarded as the nearest thing to a "supreme being" that exists. So here is the primary issue: Which group contains the truth? One of them, by absolute logical requirements, must do so. (If perfection is meaningless, this only makes case seven, that is, group three, true a priori.) When we know the answer to this question, we shall at least know whether or not the usual view of God ("usual" in philosophy and theology, perhaps not really usual in religion) is sound, and whether or not atheism or something close to it is sound, or whether, finally, the truth lies in a less explored region, the second group.

It must in all this discussion be understood that certain doubtful or trivial meanings of "perfect" or "unsurpassable" are excluded (merely to save time and energy), such as that a squirrel is perfect if it has all that is demanded by the concept (whose concept?) of a squirrel, or that a nail is as good as any other could be if it holds the building together as long and as well as is wanted. Such merely subjective or merely instrumental perfection is not what is meant by the perfection of God. It is not for this or that special purpose or point of view that God is unsurpassable. Rather it is his purpose and point of view themselves which are thought to be unsurpassable and the very standard of all other purposes or perspectives. Everything is good merely for something except persons, or at least sentient beings, but these are good in themselves. God (if he be an individual) must be at least sentient, or he is anything but unsurpassable.

These things being understood, it follows that one, and only one, of the following propositions must be true:

1. There is a being in all respects absolutely perfect or unsurpassable, in no way and in no respect surpassable or perfectible. (Theism of the First Type; absolutism, Thomism, most European theology prior to 1880.)

2. There is no being in all respects absolutely perfect; but there is a being in some respect or respects thus perfect, and in some respect or respects not so, in some respects surpassable, whether by self or others being left open. Thus it is not excluded that the being may be relatively perfect in all the respects in which it is not absolutely perfect. (Theism of the Second Type; much contemporary Protestant theology, doctrines of a "finite-infinite" or perfect-perfectible God.)

3. There is no being in arty respect absolutely perfect; all beings are in all respects surpassable by something conceivable, perhaps by others or perhaps by themselves in another state. (Doctrines of a merely finite God, polytheism in some forms, atheism.)

This division is exclusive and exhaustive. To prove any two of these propositions false is to establish the truth of the remaining proposition; there can be no "higher synthesis" which combines the truth of any two or of all three of them, except as this synthesis amounts to accepting some one of the three as it stands and contradicting some part of each of the other two; that is, one of the three must be the higher synthesis. One may subdivide the three cases, but one cannot evade the necessity for rejecting some two and affirming some one of them as a whole, or else giving up the theistic question, the latter option being not an additional objective possibility but merely a subjective attitude toward the three possibilities. Of course one might say that there are two Gods, one corresponding to the first proposition, the other to the second proposition without the initial negative clause. But this would merely be a special case under Proposition One, and would have importance only if Proposition One is acceptable as it stands and Proposition Two false as it stands. After we have decided, if we do so decide, that there is one God wholly, partially, or not at all absolutely perfect, it will then be time enough to ask if there is also another God with another of the three characteristics.

Would it not be satisfying if the debate between atheism and theism turned out to have been so stubborn because the truth was in neither, as traditionally conceived, but in a middle ground not by any means a weak compromise between them but a clear-cut alternative as definite and legitimate, formally regarded, as any other? Without pretending here to anything like conclusiveness, I will give some reasons for taking this possibility seriously.

First of all, what does religion (not theology) say as to the three groups? Suppose the usual religious ideas of omniscience, omnipotence, and holiness or supreme righteousness be accepted. This seems to mean that God is absolutely perfect in knowledge, power, and ethical goodness. Does it follow that he is absolutely perfect in all respects? What about happiness or bliss? Surely religion is not, at any rate, so emphatic here. Is not God displeased by sin, and so something less than purely happy in beholding it? Does he not love us and therefore sympathize with our sufferings, wish that they might be removed? Do we not wish to "serve" God, carry out his purposes, contribute to his life somehow? All this must be explained as extremely misleading, if not indefensible, if God enjoys absolute bliss in eternity. But, you say, would not perfect power, wisdom, and goodness insure perfect bliss? Not at all, I answer with all the conviction I can feel about anything. To be happy is not a mere function of these three variables. For to know all that exists is not to know all that might exist, except as potentialities, and if potentialities are as good as actualities, then let us all cease to exist and be done with it. It is not even true that the omniscient must know details of the future, unless it can be proved, against Bergson, Whitehead, Peirce, James, and many others, that the future has any details to know.³ (Of course it will be detailed, but this does not imply that it has detailed will-be’s as parts of itself now. . .)

Thus there is no reason why perfect knowledge could not change, grow in content, provided it changed only as its objects changed, and added as new items to its knowledge only things that were not in being, not there to know, previously. Again, to have perfect power over all individuals is not to have all power in such fashion as to leave the other individuals none. For to be individuals and to have some power are two aspects of the same thing. So even the greatest possible power (and that by definition is "perfect" power) over individuals cannot leave them powerless, and hence even perfect power must leave something to others to decide. And if one loves these others, and their decisions bring conflict and suffering, how can one, as loving toward them, escape a share in this sorrow? We know nothing of the nature of benevolence in ourselves if it is not a sharing, at least imaginative, in the interests of others, so that the partial defeat of these interests becomes in a real sense a partial defeat for us. Thus, perfect goodness is not a sufficient condition of all possible bliss. Rather, the good person suffers more than the bad at the spectacle of the badness and suffering of others. The dilemma appears final: either value is social, and then its perfection cannot be wholly within the power of any one being, even God; or it is not social at all, and then the saying, "God is love," is an error. It may be said, however, that I have confused love with desire. I reply, Love is desire for the good of others, ideally all others, or I have yet to be told what it is.

So religion does not decide clearly in favor of group one, and seems rather to support group two. God is absolutely perfect (and in so far "without shadow of turning") in those things that depend by their nature upon one’s own excellence alone. There is, for instance, nothing in the idea of knowledge to imply that God could not know all that goes on in the bad man as well as in the good; but if he equally derives (or equally does not derive) bliss from the two, so much the worse for his alleged goodness!

Inspection of the table of seven cases reveals also interesting implications for philosophy. If there is a being corresponding to case one, then there is a being totally exempt from the possibility of decrease or increase in value, hence of change in any significant sense. In such a being time is not, or at least is not time, which implies certain well known philosophical paradoxes. If, on the other hand, there is no being corresponding to any of the cases except those in the third group, if, that is, even the highest being is in all respects without absolute unsurpassability, then there is no individual being not capable of change (at least improvement) in any and every respect whatever; and in that case there is no enduring individual whose identity through all time is assured, for self-identity is incompatible with "change in all respects whatever." This threatens the intelligibility of time from the opposite point of view, for time must have some identity as well as differences. And it threatens religion, for the service of a God whose permanence is not assured fails to add anything essential to the service of men; and, moreover, the perfection of God is the heart of religious thought and feeling.

From another point of view one may reach the same result. Absolute and relative are polar concepts and seem to require each other, yet only group two makes this polarity affect the nature of the basic substance or individual. In religious terms, God, according to group two, is not just the creator opposed to the creatures, nor is he just another creature, but he is the creator-with-the-creatures, his reality is not in all respects as it would be did the creatures not exist. . .

As among the three cases under group two, it might appear that case three (ARI) is the most promising of all, since it alone combines all three fundamental categories (surpassability by nothing, surpassability by self only, surpassability by others than self). But the third category is in a sense derivative. God can very well embrace surpass-ability by others, but as his property only insofar as it is that of relative beings united to him by virtue of his relative aspect. Thus if x comes to be surpassed by y, then God in his total value, as first including the value of x and then the value of y, will surpass himself in a manner which will be the reality of the x and y relation as enjoyed by him. But if God were incapable even of self-surpassing, then no surpassing could contribute anything whatever to his value or mean anything to him, for to him there would be no more or less but just sheer value.

On the other hand, as between cases two and four (AR and Al), the apparent choice is in favor of two. For Al implies that a being consists exclusively of an absolute fixed perfection plus a purely changeable and surpassable imperfection; or in other words, insofar as the being changed at all there would be no ultimate limit of any sort to this change, and no guarantee that the being which in some respects was absolutely perfect would remain even superior to others in his non-absolute aspects. Even supposing that two such pure opposites could constitute one individual or entity, this entity seems to have little to do with anything that has been meant by God.

Thus we have some reason for suspecting that the second case, AR, the farthest removed from atheism or pure relativism, the closest to the theological tradition, is the truth of the whole question. Since it is five steps away from atheism out of a possible six, lovers of the letter of orthodoxy who might feel inclined to attack case two as little better than atheism, or as a blasphemous or at best a crudely inept doctrine, might pause, before indulging in such judgment, long enough to consider — and I am confident they will not have done so before — what the five steps really mean. They mean, in fact, that most of traditional theology is acceptable to AR theorists as a description of one aspect of God, the A aspect. Yet since, on the other hand, the single step separating case two from the older theory involves the entire difference between admitting and not admitting real change, growth, possibility of profit. suffering, true sociality, as qualities of the divine, along with radical differences (as we shall see) in the meanings ascribed to creation, the universe, human freedom, and in the arguments for the existence of God, those inclined to think that any view that is intimately connected with theological traditions must have been disposed of by this time should also beware lest they commit a non sequitur. And finally, those who think that the modern experiments with a "finite" God have proved abortive might take heed of the radical ambiguity of all such phrases, and of the logical independence of case two from all of the four or five doctrines which could most reasonably be meant by them.

It is not even to be assumed that case one, at the opposite extreme seemingly from atheism, is really in every sense "farther" from it than is case two. For the "line’ connecting the seven cases may be self-returning, if more than one dimension be involved. And this condition is here fulfilled. Case one makes God no more superior than does case two in the dimensions covered by A in AR, and it makes him infinitely less perfect in the R dimension, if any, for these are such as to imply change, self-transcendence, for their value — as, for instance, does novelty as a dimension of value. Also, as we have seen, trying to treat these R dimensions under A might destroy even the dimensions to which A is appropriate. So the God of A might really and consistently have even less perfection than the human race, or whatever the atheist regards with such reverence as he may feel. Hume’s Dialogues (Part IV) are one of the earliest expressions of insight into this meeting of extremes.

The formal analysis of perfection makes evident the absurdity of supposing the theistic question to be a mere product of superstition or of some "complex." The notions which define perfection are logically inevitable in philosophy. Either these notions admit consistent combination as required for the definition of perfection (in one or more of the six senses) or they do not. This depends solely upon the meanings of "greater," "none," and "possible." Hence if we do not know whether or not perfection is conceivable, and in what sense or senses, we do not know what we mean by concepts than which none could be more elementary in philosophy. .

Exact thinking, it is rather generally agreed among those noted for it, is mathematical, or rather has at least a mathematical aspect, however complex or simple. (In very simple cases, mathematical symbols may scarcely be required.) It will have been observed that the formally possible modes of unsurpassability are simply the mathematically possible combinations of the ideas required to render "unsurpassable" univocal in meaning. This is an application of mathematics to the greatest of human problems, an application not less legitimate or important because so elementary and simple that it seems prodigious talent must have been required, and certainly was in fact expended, to overlook it for so many centuries. As in all cases of applied mathematics, truth cannot be certified by the mathematics alone. What can be certified is the definiteness and completeness of the possibilities among which the truth, so far as statable through the concepts initially proposed, must lie. There is no other way whatever of insuring that the truth does lie between given alternatives, rather than in some alternative not even consciously considered. Those who may fear that the use of exact formal concepts must somehow be hostile to religion will insofar be true enemies of knowledge as well as doubtful friends of religion. But just as Bradley affected to quarrel with arithmetic, so we should expect that some will dislike the attempt to arithmetize theology. Exact thought has its enemies.

It will be noted that unsurpassability is verbally a pure negative. It can be correlated with a positive idea by the notion of totality. If a being has "all" the values that exist, then it is in all respects unsurpassed by anything actual. If it has all the values that are possible, then it is unsurpassable by anything possible. But if all values are not "compossible," cannot all coexist, as seems an almost obvious truth, then a purely final or static perfection possessing all possible values is impossible. We must then conceive perfection as partly dynamic, in some such manner as follows:

A being may have a relation to all actual values which, as a relation, has all the value possible, or as much value as possible, in view of the relata (the values given as actual), and the being may have a relation to all possible values as such which, as a relation to possibilities, could not be superior. Such a highest possible relation to actual and possible value might consist in this: that all possible values would, if and when actualized, belong to the being in question, that is, the being would always be unsurpassable, except by itself as it actualized more and more of the possibilities confronting it. Yet as possessing thus at all times the highest possible abstract type of relation to actuality and possibility the being would, in one aspect of itself, enjoy absolute or static perfection, be not only unrivaled but even incapable of improvement. All that is necessary to reconcile this with the religious idea is to show that such absolutes as omnipotence or omniscience or perfect righteousness or loving-kindness are abstract relational absolutes in the manner just indicated, and thus not only compatible with but inseparable from a qualitative, concrete aspect of perfection which is dynamic, since it involves inexhaustible possibilities for achievement. Is it not almost obvious, again, that the religious terms mentioned are abstract and relational precisely in the manner outlined?

One might try to make perfection positive in another way, by using the notion of surpassing all things rather than of being surpassed by none. But the reader will, I think, if he experiments with this idea, find that it leads to the same result. The importance of assuring a positive content for perfection is that otherwise one cannot well deny the contention of atheism that the word God is merely a word for what is left when we deny all that we know; that is, it represents what we know when we know nothing. This "negative theology" has often been praised, on the ground that all our knowledge is so inadequate to God that we must indeed negate it to arrive at God. But why not to arrive at non-being? Some positive content to the former idea there must be to distinguish it from the latter, and why not the utmost positive content, infinite, indeed? Surely a little dose of positivity will not suffice here. And the dilemma remains, even in the negative theology, that either all value is compossible — which seems certainly untrue, for values conflict — or else God must fail to possess some values which yet are possible — and how then can he be incapable of growth in value? Possibilities which to God represented no possible achievements would be the same to him as no possibilities. True, one can recognize values for others, say their joys, without fully possessing or expecting to possess these as one’s own, but what one cannot do is to fail in such a case to derive at least some value from the joys through the act of recognition itself, and precisely the most perfect mind would derive most from the satisfactions of others. It is the imperfection of man that compels him to admit that some of the joy which he wishes others to possess may when it comes contribute nothing to him, since he may be absent, dead, or somehow cut off from participation in the joy. Only the perfect can participate perfectly, gain for himself the entire sum of all actual gains.

If all values are compossible, and are all actual in God, then it is meaningless to say that some values are only possible. Possibility in that case ceases to have any distinctive meaning. Even if you say that God has not the actuality of what for us are possible values but rather a value above all our possibilities, you are only saying that what we call possibility is nothing from the ultimate standpoint. It is at least a serious thing to make the idea of God the destruction of a category without which it is doubtful that we can think at all.

The question is sometimes asked, Is God a concrete individual or is he an abstraction? If there is anything in the ontological argument, it may be that God must be concrete. For that argument may perhaps amount to this, that perfection is conceivable only as the property of an existing individual, and not of merely possible individuals (whereas we may conceive the nature of Mr. Micawber, for example, as not in fact the nature of an existing man). But even if we grant that God is an abstraction or a Platonic form or something somehow superindividual, still this does not obviate our trichotomy of doctrines. . . . The form is in all respects, in some respects, or in none an absolute ideal, the ideal of an unsurpassable maximum. The question then is, Are the dimensions of value alike in admitting, or in not admitting, an upper limit, or are there some which do and some which do not and which yet must apply to all things having value?

Our classification of doctrines depends only upon the four following assumptions:

p. There is a difference between actual and possible (or conceivable) things.

q. There may be a difference between actual and possible states of an individual. (Not that God is assumed to be an individual in this sense, but that it is not assumed that he is not, in the statement of the classification, whose purpose is to state, not to answer, controversial questions.)

r. It is meaningful to say that one thing is higher or better than, or superior to (or has more of some variable property not a mere deficiency than), another; but this meaning is not simply univocal, since x may be better than y in one respect, say in ethical goodness, and not better in another, say in happiness. Thus "better than" is multi-dimensional. (The doctrine of the tradition that God is not simply better than other even possible beings, but is better than goodness itself, better than "best," since he transcends the concept of goodness altogether, does not alter the necessity that he be better-than-best in some, in none, or in all dimensions of value; or negatively, that he be surpassable in all, some, or no dimensions. The tradition spoken of clearly elected the first of the three formal cases, making God unsurpassable by anything conceivable, even by potential states of himself.)

s. The notions of "all," "some," and "none" exhaust the possible divisions of a plurality, hence of a plurality of respects of higher and lower. (Logicians distinguish between "all" and "every," but this seems of no importance here.)

These assumptions (except the last, which is clearly self-evident) are not posited absolutely. It may, you may believe, turn out that actual and possible coincide, or that the different dimensions of value or superiority are really one. The point is, we must not assume this at the outset. What we certainly must assume at the outset is that the question of such distinctions requires discussion, and that therefore every type of doctrine implied as formally possible if the distinctions are genuine must be given full and fair hearing. If two views formally distinguished turn out to be the same (since some alleged distinction separating them proves equal to zero), then that will be the conclusion reached; but it must be a conclusion, and not in any sense a formal premise, of the argumentation. There can be no harm in setting a terminological locus for alleged distinctions, admitting that they may assume every value of significance from zero to infinity; but there is very definite harm in depriving apparent distinctions of terminological and systematic locus, since their value is then determined as zero by fiat. Now the distinctions between "superior to actuality" and "superior even to possibility," or between "superior to other possible individuals" and to "other possible states of oneself" (as an individual identical in spite of changes or alternate possible states), or again, between "superior in all," "in some," or "in no" respects of value — these distinctions are urged upon us by universal experience and common-sense modes of thought. They may be overruled in the outcome, they can never validly be overruled before the outcome, of technical procedure. And we have painfully learned (all but one or two groups of philosophers) that the way to evaluate ideas is to deduce their consequences and compare these with the relevant data of experience. So we have no rightful alternative to the systematic development of the consequences of the distinctions mentioned. The discussion of the resulting doctrinal classifications is the bottleneck through which alone we can arrive, if ever, at a rational treatment of the theistic question.

This question can, it is true, be put in other initial terms than those we have used. For instance, it can be put in terms of causality. Has the world a cause, or is it self-sufficient? But this formulation is not precise. It suggests that God is nothing but causation, and the world nothing but effect; in other words, that God is in no sense affected by other individuals, and the world in no sense causal in relation to God. But the idea of God in its common-sense or religious meaning may not require this. God is of course the supreme power in existence, the causal influence superior to all others. It remains to be seen, however, whether superiority of power implies a purely one-way causal action, an action without reaction or interaction. That is a basic technical question, not to be decided near the beginning of discussion but toward the end. Perhaps the supreme action is also, necessarily, the supreme interaction. Nor can words like "creator" and "creation" dispose of the matter. Religion is not prima facie committed on such technicalities as the relation of creativity to various causal concepts.

In terms of causality there are, rather, three formal possibilities, corresponding to, indeed coinciding with, our basic trichotomy. The highest cause may be (1) in every sense or aspect "uncaused," in no sense or aspect the effect of anything else; or it may be (2) in some aspects uncaused, and in others causally influenced, but its manner of both acting and receiving influences may be the highest conceivable, hence absolutely "perfect," although even so its whole being may not in every sense be perfect, because the influences as coming from other causes, say human beings, may be less admirable than they might be; or the supreme cause may be (3) in no sense or aspect uncaused, independent of other powers, hence in no way wholly exempt from the imperfections of the latter. . . .

It makes no difference what concepts are used, whether "self-existent," "necessary being," "unity," "final cause," or what you will to describe the divine individuality; there are always three formally possible cases (though the boundaries between them could be variously located, and they can be subdivided) among which choice must be made openly and carefully, not surreptitiously nor by a short and easy appeal to self-evidence. A being may, for instance, be necessary in all its aspects, or not in all but in some, or, finally, in none. So with all the other concepts mentioned above. Nothing can result but endless debate (and bad feeling) from the attempt to short-cut the exploration of an irreducibly triadic situation. Dyadic formulations of the theistic problem are question-begging through and through. . . .

Naturally any view which ascribes ethical perfection and yet the "greatest possible power" to God must face the problem of evil. In its appeal to the imagination this problem will no doubt always be the most troublesome one in theology. But in pure logic it is not true that there is sheer contradiction between the joint admission of divine perfection of goodness and divine perfection of power, on the one hand, and the fact of real evil on the other, for the simple reason that the greatest possible power (which by definition is "perfect" power) may not be the same as "all the power that exists united into one individual power." For such union of "all" power may be impossible. Had God "all the power there is," he must be responsible for all that happens. But why assume that all real power could possibly belong to one individual? If it could not — and there is ground for this negative — then even the perfect or (by definition) greatest possible power is not all-power. Omnipotence (alas, our only word for perfection of power!) is power to the highest degree possible and over all that exists, it is "all" the power that could be exercised by any one individual over "all" that is; but it remains to be shown how much power could be exercised in this fashion. The minimal solution of the problem of evil is to affirm the necessity of a division of powers, hence of responsibilities, as binding even upon a maximal power. But this solution seems to imply the passivity of the supreme power, and hence not to be available to first-type theists.

Undoubtedly, "ethical" needs careful defining, but roughly it means action issuing from the fullest realization available to the individual of all the interests affected by the action. It does not necessarily mean observing the rules or codes recognized in any human society, except insofar as these represent the attempt of that society to make actions express the nearest thing to full realization of affected interests which is possible to the average human being. Being ethical does not mean never injuring anyone; for the interests of others may require such injury. Still less does being ethical mean never permitting any agency to bring injury to anyone; for not permitting this might be possible — owing to the division of power — only at the cost of greater injury through interference with other powers. Being ethical means acting from love; but love means realization in oneself of the desires and experiences of others, so that one who loves can insofar inflict suffering only by undergoing this suffering himself, willingly and fully. Those who think God cannot mean well toward us because he "sends" us suffering can prove their point only by showing that there is a way to run the universe, compatible with the existence of other real powers than just the supreme power, which would be more fully in accord with the totality of interests, or by showing that God sends us the suffering while himself remaining simply outside it, in the enjoyment of sheer bliss. Theologians themselves (first type) seem generally to have made a present of the latter notion to atheists; but the former view has its plausibility for all of us. I wish only to say here that I think neither is put beyond reasonable doubt by metaphysical necessity or empirical facts. It is poor method to try to estimate facts, especially such as are hard to measure with any accuracy, without careful survey of the logical structure of the ideas we bring to bear upon these facts. Therefore the facts of evil are not sufficient to justify dismissal of theology prior to the adequate exploration of its three main formal possibilities. Facts will never render decisions between ill-conceived alternatives; and the meaning of such terms as omnipotence or goodness depends in second-type theism upon a number of conceptions which have not been clearly considered in the classic discussions (such as the marvelous one in Hume’s Dialogues) of the relations of such terms to the facts of evil.

One way of trying to escape a decision among the three possible views concerning God as a perfect being would be to say that perfection as "that than which nothing higher or better in a given respect is conceivable" is a meaningless concept, itself inconceivable. This, however, besides seeming tolerably dogmatic, would only be to say that Proposition Three is true by necessity; for if a predicate is nonsense, then of course nothing exists having that predicate. Hence no form of positivism can provide an evasion of the decision to be made.⁴ Nor can any other doctrine do so. What we have is a non-controversial statement of what the theistic controversy is. In general, I believe, all stubborn controversies in philosophy have involved questions the very existence of which as such is itself controversial, because they have not been formulated in neutral terms, terms that avoid arbitrarily limiting the prima facie possibilities.

In particular, most philosophico-theological controversies have amounted to one of the following procedures:

A. To considering reasons for preferring one or the other of Propositions One and Three, or more probably, some special variety of One to some variety of Three;

B. To considering reasons for preferring some one variety of One (such as "theism" or "absolutism") to some other variety of One (such as "pantheism" or "deism").

A is bound, sooner or later, to involve the fallacy of inferring the truth of One from the falsity of Three, or vice versa; whereas it is formally possible, and should be held really possible, until the contrary has been shown, that both One and Three are false because Two is true. The fallacy is bound to occur so long as Two is neglected, for the reason that men do not adopt a philosophy because its proofs are beyond question and its conclusions completely satisfactory — this being never the case — but because its proofs seem to them stronger and its conclusions more satisfactory than would be true of what they regard as the alternative. It is a question of preference, not of absolute unclear evidence and perfect understanding. In so far as this is the case, almost everything depends upon the adequacy of the philosopher’s survey of the possibilities. Now there is no more rigorous trichotomy than that of "all, some, none"; hence the question, Is God absolutely perfect in all, in some, or in no respects? is as rigorous a division of the theological problem as can be given if any use at all is to be made of the idea of perfection — and what theology has avoided its use? Moreover, if all the formal possibilities are not controlled, we not only run the risk of fallaciously inferring the truth of one view from the difficulties of some only of its possible rivals, hut also we run the risk of trying to answer a perhaps meaningless question, namely, Which of two falsehoods (or absurdities) is more false? The falsehoods may be extremes (and One and Three are clearly such), and hence one may be as false as the other, by any objective standard. In that case, the choice between them will be on subjective and variable grounds, and no agreement is to be anticipated. If then, under these circumstances, complete agreement is not reached, it does not follow that agreement could not be at least greatly increased by the accurate, exhaustive statement of the doctrines open to us, arranged in a reasonably small number of exclusive groups or types.

B is an attempt to decide upon the details of a type of theory whose admissibility as a type has not been shown, owing to the role of the fallacy mentioned (which is implicit both in traditional proofs for God’s existence and in atheistic criticisms of these proofs). This does not mean that such discussions have accomplished nothing, but it does mean that no exact and reliable estimate of what they have accomplished (though it is, I believe, a great deal) is possible until we have granted full "belligerent rights" to second-type theism, as a no less qualified contender than either of the others. True, this type of theism has already had a good many defenders; but taking philosophers as a whole and theologians as a whole it is still far from true that the theological problem is seen in terms of its fundamental trichotomy, systematically investigated. . .

Our basic trichotomy of doctrines may be put in still another way, which also gives a clue as to the possible validity of the neglected second type. If we define a "closed" dimension of value as one of which there can exist a supreme or maximal case, and an "open" dimension as one of which no supreme case is possible, then one of three things is true: all dimensions of value are closed, some dimensions are closed and some are open, or none are closed and all are open. It is indeed not formally evident that the first proposition defines first-type theism; for we have not specified or shown that the maximal case of the different dimensions must be found in the same real individual. But at least it is clear that if, and only if, the first of the dimensional propositions is true, first-type theism may be true; and that if the second dimensional proposition is true, second-type theism may be true, for then there may be a real case of perfection on some dimension which will not be a case of perfection upon all, because — by the assumptions not all admit of perfection. (If the ontological argument were shown to be valid, the may be true" would in both cases imply "is true.")

Now, is it particularly obvious that all dimensions of value must be closed dimensions, assuming some of them are? Consider the dimensions of goodness, knowledge, power, and duration. A being may perhaps be the maximal case of goodness if he guides his action by concern for all the interests affected by his actions. This "all" is the universe (up to the present, at least) so far as it contains values. Or, a being may be omniscient if he knows all there is to know: that is, again, the cosmos as a totality. A being may, similarly, be the maximal possible power if he controls all that exists to the greatest extent possible, that is, to the extent which is compatible with the measure of independence, if any, constitutive of the things controlled. Finally, a being may have maximal duration by being ungenerated and immortal, by enduring throughout all time. So far, our dimensions seem to admit of maxima as at least conceivable.

But there are other dimensions of value. What could be meant by maximal happiness, or beauty, or "intensity" of joy, or variety, "the spice of life"? A being may enjoy all that exists, but perhaps he longs for what does not exist; or perhaps some of what exists is not altogether enjoyable (such as the sufferings of other sentient beings). Oh, well, you say, but if the being has maximal power, he can produce such beings as he wishes to enjoy. But there is social enjoyment, and this by definition depends partly on the self-determinations of the beings enjoyed. This cannot possibly be wholly coerced by any one term of the social relation, hence not even by the maximal "possible" power. The only escape at this point is to take shelter in the doctrine of the Trinity, which offers to furnish a social relation between persons all of whom are perfect. But still, we may ask, what in this relation is enjoyed? Is it "unity in variety," as seems to be the case with us? Supposing that variety in God is really compatible with his alleged simplicity, we still have to ask, What is meant by maximal variety? Is it that all possibilities are actualized in one actual state? But there are mutually incompatible alternatives (or there is no such thing as logic, or aesthetics). Besides, if all potentiality is also actuality in God, then the distinction between potential and actual must really be an anthropomorphic illusion, invisible from his point of view. At any rate, enjoyment varies as to intensity, and what can be meant by "all possible intensity," or "absolute intensity"?

Of course one could argue that an open dimension involves an infinite regress, and is therefore impossible. But . . . the infinite regress in question is an example of the "non-vicious" type of regress, since it concerns possibilities, and these not (on one view of potentiality) as a definite multitude, whose number is infinite, but as a continuum, which in the words of Peirce is "beyond all multitude," as God was formerly described as being; and indeed, as we shall see, the continuum of possibilities is one aspect of God which may be truly so described. It has also been argued that the maximal case is required as the standard or measure for all cases (Plato). But it may be that the maximal case on the closed dimensions would suffice to furnish the standard for the open ones, that, e.g., perfection of knowledge and goodness is in some sense the "measure" of degrees of happiness, even though the latter cannot be absolutely but only relatively perfect (R but not A).

Let us return to our conceivably closed dimensions and ask if they are not really ambiguous, not really in one sense necessarily open as well as, in another sense, capable of upper limits. To "know all that exists" is, in one sense, to have perfect knowledge, it is literal omniscience (provided possibilities are also known as such, as a special class of existences or, at least, of realities). But perhaps some of what exists is not as well worth knowing as some other things would have been had they existed. This implies no error or ignorance on the part of the knower, but it does imply the possibility of an increase in the aesthetic satisfaction derived from his knowledge, should a more varied or more harmonious world come into existence and be known. Again, one might deal justly and mercifully with all of one’s world, and still be glad should this world itself improve in some way. The justice or mercy will not be improved from the ethical standpoint, but the just and merciful one will rejoice and gain in total satisfaction should the individuals being dealt with increase in goodness or happiness. Similarly, maximal power over a good world would not be so good as maximal power over a better one, though in both cases it would be as much power as is compatible with the world to be controlled; that is, in both cases it would be maximal simply as power, though not as total value realized by the one having the power.

True, if (as we shall later see reason to question) maximal power means power to create a beginning of finite existence in time, then it would seem that God could have started with as good a world as he chose. But a "best world" may be meaningless. And besides, the very next moment he would begin to confront the results of the choices, the exercises of power, granted to the creatures, and from then on his actual state, as constituting his knowledge, goodness, and power relations, would be as we have described it.

Nor does it help to argue that since God is timeless he knows and enjoys in advance all that the world ever will become. For he cannot enjoy all that the world ever could become as much as he would if it actually became it; for example, he cannot enjoy all the good deeds men might have performed as much as he would have, had the good deeds been performed. At least, this must be so if any vestige is to remain of religious ethics, and even perhaps of good sense. No more does it help to suggest that God’s value is wholly independent of his relations to the world, whether of knowledge or of will, for this only means that the particular characters of the objects of his knowledge, or the results of his willing, are to him totally insignificant, which is psychologically monstrous and is religiously appalling as well..

Thus we have every reason to take seriously, as the tradition has plainly not done, the hypothesis (at present merely that) of open dimensions of value, even for the perfect one. Let us remember that number is incapable of a maximum, that in whatever sense God may be "beyond number," still number can hardly be in every sense without value to him — or at any rate, variety can hardly be, and there is no more reason to speak of maximal variety than of maximal number. If, however, variety is said not to be a value for God, then one asks, Why a creation at all? Why should he add to his own perfection the contrast of the purely inferior creatures, unless contrast as such is valuable? And then, how can there be a maximum of contrast? It is no use to say that God creates the creatures out of generosity or love; for if he loves the valueless, so much the worse for his love, and what but the value of contrast can the creatures add to existence? Admittedly, they do not add "unity"!

Here then is a theology that either means nothing certainly identifiable (without supernatural grace or high genius in the art of reconnecting with experience concepts carefully divested of relation to it) or else means that the world might exactly as well not have existed, or as well have existed with far more evil or less good in it than it actually presents. In short, we have the view that the world, including the theologian, is strictly valueless to God, an absolute nullity from the standpoint of ultimate truth. I submit that this is a theology to be accepted, if at all, only after all other possibilities have been carefully considered and found hopelessly untenable. If a man denies this, I only say that I scarcely believe he is thinking about what he is saying. And the writings of those who apparently do deny it show little enough evidence of thought on this aspect of the question. The very question seems, by a near-miracle of persistent looking the other way, to be passed over. Is this merely the "method of tenacity" or is there a more generous explanation?

The theological views of Philo, Plotinus, Augustine, St. Thomas, Spinoza, Leibniz, Kant, Schleiermacher, Royce, the Hindu Sankara, present differences that are striking enough, but all of them agree, or fail clearly to deny, that God is a being "absolutely infinite" (Spinoza’s phrase) or every way complete and perfect, and there seems little rational place for significant variations of opinion in a doctrine so completely determined as the doctrine of complete perfection. If, nevertheless, historically endless disputes and radical disagreements over the interpretation of the doctrine have in fact arisen, this is one piece of evidence that there is probably something wrong, perhaps self-contradictory, in the basic idea. On the other hand, the proposition that God is both perfect and perfectible, or both statically and dynamically perfect, unsurpassable, tells us prima facie nothing as to the respects in which he is the one and those in which he is the other. Here the necessity for exploring various interpretations is obvious. The exploration, however, was left largely to the present century. The opportunity this represents will not be brushed aside too hastily by anyone trying to be scientific in philosophy, whatever his religious or philosophical tenets. . .

Controversies between theism and atheism have generally leaped over one of the three basic possibilities. People have rejected theism because they held untenable the idea of a mind not subject to change or to interaction with other beings, or a mind omnipotent in the sense that its power was all the power in existence, or a mind having precise knowledge of details of the future (or of all times from the standpoint of eternity), or a mind creating a first state of the cosmos at a finite time in the past, or knowing all suffering although it did not itself suffer, or an all-embracing mind which in no sense could be identified with the universe, or one which could in every sense be identified with it. These and other difficulties, which may be called the absolutistic paradoxes, have force against Proposition One, but are not pertinent objections to Proposition Two. But, on the other hand, it is quite unjustified for theists to hold that we must tolerate or swallow the paradoxes or explain them away (by feats of ingenuity so subtle, and verbal methods so remote from intuitive insight or definite logical structures, that only deity could know with any assurance what was taking place), giving as justification the claim that the alternative position of atheism is even more paradoxical (lacking, it may be urged, any principle of cosmic explanation at all). The fallacy of such reasoning is clear once we see that atheism is not the only alternative to the assumptions which generate the absolutistic paradoxes. Nor, as we have seen, is the remaining alternative pantheism in any traditionally considered sense.

It might be objected to our trichotomy that there are many degrees of "some" between none and all, and that consequently nothing very definite is described by Proposition Two. However, the "some" refers to dimensions of value as significant in describing God’s perfection or perfectibility, and these dimensions are so interrelated that if we could come to a decision in regard to a very few of them the decision as to the others would probably follow. Also we could agree to classify under the third proposition all views which ascribe no more perfection to the gods than did the Greeks to their Olympians, whose only point of absoluteness seems to have been their immortality. (Any finite god held to be ungenerated as well as deathless ought perhaps to be held a minimal case of the finite-infinite God of second-type theism.)

It is of some interest to note that atheism and primitive polytheism are of the same basic type. This does not prove that if polytheism is false, atheism must be; for they are subalternatives within their type. But it does suggest that the radical falsity of primitive religious ideas as they stand is not an argument for atheism, as it is rather commonly held to be. Also the fact that atheism is at least as old (as a philosophy) as theism of the second type (it was much more familiar to Plato, for instance) suggests that there is nothing philosophically very advanced or sophisticated about atheistic doctrine as such. A really clear expression even of first-type theism is apparently indefinitely later than atheism. All of which of course proves nothing except the irrelevance of certain supposed arguments for atheism, arguments more subconscious and informal than explicit and official, but still influential.

The philosophical importance of admitting some nonabsolute aspects of God is in the resulting applicability of such categories as change, passivity, complexity, and the like, to him, and for this purpose surpassability of God, as he actually is, even if only by God himself as he could or can be, is entirely sufficient. Now though the actuality of deity is, according to second-type theism, in some respects surpassable, his individuality as potentially inclusive of other than his actual predicates may be in no respect whatever surpassable, in all dimensions though not in all senses perfect. To say this is not to commit second type theism to the view that God is an "individual." We are speaking of subalternatives which the second basic proposition admits, not of corollaries which it necessarily implies. All the proposition demands is that there be a God in some respect unsurpassable, in some other surpassable — whether self-surpassable and how, or surpassable by other entities not states of himself, or whether he has "states," being left perfectly open by the proposition. Exploration of the subalternatives may well lead to the conclusion that only one of them is really "conceivable" in the full sense (in the light of the experiential content of the ideas involved). But this again is a matter to be held in suspension until we have established some control of the relations between the basic propositions.

God, for both old and much new theology, is the being whose uniqueness consists in his unrivaled excellence, or whose amount of value defines a necessarily one-membered class (and so in a sense not a class). In some respects he is absolutely unexcelled, even by himself in another conceivable state; in all other respects he is (to state the view reached in this book) the only individual whose states or predicates are not to be excelled unless he excel them with other states or predicates of his own. To take an imperfect analogy, no one will ever be or can ever be so Wordsworthian as Wordsworth; but Wordsworth himself, if he (or someone about him) had made a different use of his free will, might perhaps have been somewhat "more himself," might have developed his individuality more than he did. And certainly, at any stage in his life, one could have said that he was the most Wordsworthian being that would ever exist, except as he himself might later become more so. God, however, is not simply more himself than any other can ever be; he and he alone is in all respects superior to any state that will ever characterize any individual unless it characterize him. He is the greatest conceivable actuality, except perhaps as he himself can be conceived as greater (in another, perhaps subsequent, state, or in a state he might have had in the past, had men, say, served him more faithfully).

There is a slight ambiguity in the expression "excelled by himself only." We may ourselves in the future enjoy values which God now lacks (because they are not in being). But according to AR he will not lack them when we enjoy them, so that our self-excelling will be also (infinitely magnified) his self-excelling. Thus R means that "in no possible state of affairs can there be anything in any fashion superior to God as he is in that same state of affairs."

It will be seen that the new doctrine requires careful and somewhat elaborate distinctions, and yet, if some of its supporters are right, the doctrine is nothing at all but the analysis of the simple idea that God is "the perfectly loving individual," in all respects possessed of the properties which this idea requires, even if non-perfection in some respects be among the requirements.

That God is less than he might be (though more than anything else might be) agrees with the religious conception of the free service of God. For if we had no choice but to serve God in the fullest measure, or if we could not serve him at all, then it might be held with some plausibility that he is all that he might be. But the possibility of being freely served seems clearly to imply the possibility of lacking something that better service than may actually be given would furnish. Philosophical orthodoxy has had to finesse this point, and indeed, as I believe, has fallen into sophistry of a rather revolting kind. Really there was to be no service of God, but only a service of men through the — to them — beneficial practices of religion. Sin did no real harm whatever in the universe, since the absolute perfection which the universe involves in its cause could never be more or less than absolute. To say that sin at least harmed men is beside the point; for what harm did it do to harm men, parts of a system of reality that as a whole or in its ultimate reality was incapable of loss or gain? The world plus the absolutely infinite is no more than the latter by itself. Only from a purely race-egoistic (and illusory) point of view could the harm appear as such. Thus the motivation which is the (attempted) attitude of pure atheistic humanism was the only one philosophers could approve in religion. The idea of cosmic concern, concern for the divine values, must now at last be considered on its merits. . . .

It will be seen that the God of second-type theism is not without qualification finite, or growing, or emergent; nor, without qualification, is he the contradictory of these. The traditional distrust of simple statement, and of language as applied to the religious vision, in the new theology ceases to be an inoperative or inconsistently employed formal concession, and becomes a systematic tracing of the relativity of concepts to each other and to experience as a whole. The concepts which still function as absolute are the strictly religious and experiential ones of love and goodness. God is the Holy One, the ethical Absolute, the literally all-loving Father. In these affirmations second-type theism sees no exaggeration. It holds that the distinction between God’s ethical perfection (and hence ethical immutability) and his "aesthetic" perfectibility (and hence growth) fits the later Hebrew and other high religions (most of all what some of us would mean by Christianity) far more naturally and unambiguously than does the confusion of every perfection in the unchanging actus purus of the Scholastics (and even of Schleiermacher). Furthermore, Whitehead and others have shown that it is precisely love which must be perfect in God — and only love and what is implied by it as perfect — if either love or perfection is to serve as an explanatory concept in cosmology. . .

What has been discovered . . . is that, on one main point at least (the choice between the three propositions), religion at its best was literally and philosophically right, and theology was but a first approximation, vitiated by ambiguities or inconsistencies. In Whitehead’s cosmology — which is, in the main, simply the most fully elaborated expression of tendencies widespread in recent philosophy — all existence is "social," is "feeling of feeling," forming "societies" of interlocked experiences, and societies of societies, from electronic, almost inconceivably simple and rudimentary, societies, to the universe. In this completely social philosophy (conflict, which is not denied, being also a social relation) God is that in the cosmos whereby it is a cosmos; he is the individual case on the cosmic scale of all the ultimate categories (including those of social feeling, "subjective aim," etc.) thanks to which these categories describe a community of things, and not merely things each enclosed in unutterable privacy, irrelevant to and unordered with respect to anything else. To impute purpose to God is no dishonesty in Whitehead; for he finds no real or possible thing that is not in its degree of simplicity or complexity endowed with subjective aim. And equally, he finds nothing whose feeling and aim are without sensitivity to other feelings and aims, that is, social. Hence the cosmic individual, the cosmos as the inclusive Society of societies "with personal order" is inclusively, universally sensitive, loving, and hence decidedly not purely impassive or once for all and in all ways perfect. The sense in which conflict, as well as harmony, enters into God is just the sense to which religion refers in speaking of the grief or anger of God over our suffering or sins, the grief being symbolized by the cross. Love is not identical with harmony, though it includes a measure of it. God conflicts, however, only with what he also participates in through his sensitivity or "tenderness." If Whitehead said less than this, it is the logic of his system that would collapse, and not merely its religious applicability.

NOTES:

1. One of the earliest expressions of this attitude is to be found in Otto Pfleiderer’s Grundriss der christlichen Glaubens- und Sittenlehre (Berlin: Georg Reimer, 1888), Sections 61, 67-69, 84.

2. The "new" theology can also be called Platonic if one interprets Plato somewhat otherwise than the Neo-Platonists and most scholars have done. See Raphael Demos, The Philosophy of Plato (Charles Scribners Sons, 1939), 120-125.

3. That possibilities are real, and that the future involves open alternatives, or is indeterminate in essence, I have attempted to demonstrate in my book, Beyond Humanism, chaps. 9 and 10, and in an article, "Contingency and the New Era in Metaphysics," Journal of Philosophy, XXIX, 421ff., 457ff. Cf. Charles S. Peirce, Collected Papers (Harvard University Press, 1931-1935), Vol. VI, Book I A. For an elaborate defense of the opposite or deterministic view, see Brand Blanshard, The Nature of Thought (London: Allen & Unwin, 1939), especially Vol. II. (Blanshard virtually ignores most of what seem to me the chief arguments against determinism, but gives a fine account of the arguments which have often been thought to support it.)

4. The positivistic objections to metaphysics as such I have attempted to meet in chap. 16 of Beyond Humanism, and in "Metaphysics for Positivists," Philosophy of Science, II, 287ff. See Adventures of Ideas 147f., 159-165; Peirce, Papers, VI, 368. . .